I have a grammar that is quite ambiguous. Ambiguities aren't solved easily because the grammar is quite complex and the task of resolving ambiguity is probably not possible. Ambiguities are supposed to be solved later after establishing some context in which it can be resolved.
The biggest source of parsing failures I have because in the alignment of the matching rules a short match often prevents a long match from being found. And I wonder if I can look for the longest match with petit parser. Combining parsers with / always finds the first matching and | does not find anything if more than one rule matches.
Any advize how to do this with petit?
Norbert
Order your choices from the longest to the shortest.
If you don't know the longest match statically (from the grammar) but only from the input, you could define your own choice operator along the following lines (untested code):
LongestChoiceParser>>parseOn: aStream | start element longestEnd longestElement | start := aStream position. 1 to: parsers size do: [ :index | element := (parsers at: index) parseOn: aStream. (longestEnd isNil or: [ longestEnd < aStream position ]) ifTrue: [ longestEnd := aStream position. longestElement := element ]. aStream position: start ]. aStream position: longestEnd. ^ longestElement
Lukas
On 26 July 2011 12:25, Norbert Hartl norbert@hartl.name wrote:
I have a grammar that is quite ambiguous. Ambiguities aren't solved easily because the grammar is quite complex and the task of resolving ambiguity is probably not possible. Ambiguities are supposed to be solved later after establishing some context in which it can be resolved.
The biggest source of parsing failures I have because in the alignment of the matching rules a short match often prevents a long match from being found. And I wonder if I can look for the longest match with petit parser. Combining parsers with / always finds the first matching and | does not find anything if more than one rule matches.
Any advize how to do this with petit?
Norbert
Moose-dev mailing list Moose-dev@iam.unibe.ch https://www.iam.unibe.ch/mailman/listinfo/moose-dev
Am 26.07.2011 um 12:56 schrieb Lukas Renggli:
Order your choices from the longest to the shortest.
Yes this is the normal choice to reorder were possible.
If you don't know the longest match statically (from the grammar) but only from the input, you could define your own choice operator along the following lines (untested code):
LongestChoiceParser>>parseOn: aStream | start element longestEnd longestElement | start := aStream position. 1 to: parsers size do: [ :index | element := (parsers at: index) parseOn: aStream. (longestEnd isNil or: [ longestEnd < aStream position ]) ifTrue: [ longestEnd := aStream position. longestElement := element ]. aStream position: start ]. aStream position: longestEnd. ^ longestElement
thanks, it works. Do you plan to integrate that in petit parser? I wonder if it would be feasible to implement yet another binary selector for another parser. I use /> at the moment.
Norbert
On 26 July 2011 12:25, Norbert Hartl norbert@hartl.name wrote:
I have a grammar that is quite ambiguous. Ambiguities aren't solved easily because the grammar is quite complex and the task of resolving ambiguity is probably not possible. Ambiguities are supposed to be solved later after establishing some context in which it can be resolved.
The biggest source of parsing failures I have because in the alignment of the matching rules a short match often prevents a long match from being found. And I wonder if I can look for the longest match with petit parser. Combining parsers with / always finds the first matching and | does not find anything if more than one rule matches.
Any advize how to do this with petit?
Norbert
Moose-dev mailing list Moose-dev@iam.unibe.ch https://www.iam.unibe.ch/mailman/listinfo/moose-dev
-- Lukas Renggli www.lukas-renggli.ch
Moose-dev mailing list Moose-dev@iam.unibe.ch https://www.iam.unibe.ch/mailman/listinfo/moose-dev
thanks, it works. Do you plan to integrate that in petit parser? I wonder if it would be feasible to implement yet another binary selector for another parser. I use /> at the moment.
I've added it to PetitBeta using the proposed operator #/>.
Name: PetitBeta-lr.51 Author: lr Time: 28 July 2011, 10:13:33 am UUID: ff87b4d9-b874-47fe-b85a-ee80b2321fa0 Ancestors: PetitBeta-lr.50
- added PPLongestChoiceParser to experimental
Norbert
On 26 July 2011 12:25, Norbert Hartl norbert@hartl.name wrote:
I have a grammar that is quite ambiguous. Ambiguities aren't solved easily because the grammar is quite complex and the task of resolving ambiguity is probably not possible. Ambiguities are supposed to be solved later after establishing some context in which it can be resolved.
The biggest source of parsing failures I have because in the alignment of the matching rules a short match often prevents a long match from being found. And I wonder if I can look for the longest match with petit parser. Combining parsers with / always finds the first matching and | does not find anything if more than one rule matches.
Any advize how to do this with petit?
Norbert
Moose-dev mailing list Moose-dev@iam.unibe.ch https://www.iam.unibe.ch/mailman/listinfo/moose-dev
-- Lukas Renggli www.lukas-renggli.ch
Moose-dev mailing list Moose-dev@iam.unibe.ch https://www.iam.unibe.ch/mailman/listinfo/moose-dev
Moose-dev mailing list Moose-dev@iam.unibe.ch https://www.iam.unibe.ch/mailman/listinfo/moose-dev
Am 28.07.2011 um 10:14 schrieb Lukas Renggli:
thanks, it works. Do you plan to integrate that in petit parser? I wonder if it would be feasible to implement yet another binary selector for another parser. I use /> at the moment.
I've added it to PetitBeta using the proposed operator #/>.
Name: PetitBeta-lr.51 Author: lr Time: 28 July 2011, 10:13:33 am UUID: ff87b4d9-b874-47fe-b85a-ee80b2321fa0 Ancestors: PetitBeta-lr.50
- added PPLongestChoiceParser to experimental
thanks,
Norbert
Norbert
On 26 July 2011 12:25, Norbert Hartl norbert@hartl.name wrote:
I have a grammar that is quite ambiguous. Ambiguities aren't solved easily because the grammar is quite complex and the task of resolving ambiguity is probably not possible. Ambiguities are supposed to be solved later after establishing some context in which it can be resolved.
The biggest source of parsing failures I have because in the alignment of the matching rules a short match often prevents a long match from being found. And I wonder if I can look for the longest match with petit parser. Combining parsers with / always finds the first matching and | does not find anything if more than one rule matches.
Any advize how to do this with petit?
Norbert
Moose-dev mailing list Moose-dev@iam.unibe.ch https://www.iam.unibe.ch/mailman/listinfo/moose-dev
-- Lukas Renggli www.lukas-renggli.ch
Moose-dev mailing list Moose-dev@iam.unibe.ch https://www.iam.unibe.ch/mailman/listinfo/moose-dev
Moose-dev mailing list Moose-dev@iam.unibe.ch https://www.iam.unibe.ch/mailman/listinfo/moose-dev
-- Lukas Renggli www.lukas-renggli.ch
Moose-dev mailing list Moose-dev@iam.unibe.ch https://www.iam.unibe.ch/mailman/listinfo/moose-dev
-
Lukas Renggli -
Norbert Hartl